Hume AI launches audio-native Decision API for emotion and speaker analysis
Hume AI has launched an audio-native Decision API that answers plain-English questions about recordings, such as whether a caller sounds frustrated or how many speakers are present, returning a probability per answer. The company says the API consolidates tasks that previously required more than 10 separate models into a single call. In initial evaluations, Hume reports the model agreed with a human panel's average answer 70.7% of the time, compared with 56.5% for a single listener. No pricing, availability date, or independent verification was disclosed.
IllustrationEditorial responsibility
- No named human review is recorded for this page.
- Reports are grouped by semantic similarity and deterministic rules. Language models may assist titles, summaries, translation and cross-source analysis; the page reads the event directly, while its address stays stable when the title changes.
- Summary covers the current reports
Cross-source coverage
Reporting timeline
Hume AI Claims New Audio Decision API Replaces 10+ Models With One Call
A post amplified by DataChaz states that Hume AI has introduced a new Audio Decision API that it says consolidates a workflow previously requiring more than 10 models into a single API call. The post frames the announcement as a major simplification, emphasizing the reduction from many models to one call. The claim is attributed to Hume AI and is presented without independent verification in the source item. The post also begins to reference speaker counting but the text is truncated, so no further technical details or performance figures are provided. The item is a product announcement claim rather than a full technical disclosure.
Read sourceHume AI launches audio-native Decision API that answers questions about recordings
Hume AI has launched the Hume Decision API, an audio-native model designed to answer questions about a recording by analyzing the audio itself. Users can write their own questions and answer options, such as whether a caller sounds frustrated, and receive an estimate of how human listeners would respond. In initial tests, the model agreed with the panel's average answer 70.7% of the time, compared with 56.5% for a single listener. The launch positions the API as a tool for gauging human-perceived qualities of audio, with the company reporting benchmark results against individual human listeners.
Read sourceHume AI Launches Audio Decision API Consolidating Multiple Model Tasks Into One Call
Hume AI (@hume_ai) announced a new Audio Decision API that it says replaces workflows previously requiring more than 10 separate models with a single API call. According to the post, the API can perform multiple analyses on the same recording, including counting speakers, detecting laughter, and assessing whether a caller sounds frustrated. The company frames the product as enabling multiple questions to be answered from one recording through one API call, rather than running separate models for each task. The announcement was shared on X by DataChaz. No pricing, availability date, benchmark results, or independent verification were provided in the post.
Read sourceShow 2 older updatesHide older updates
Hume AI Launches Audio Decision API for Speech and Emotion Analysis
Hume AI announced the release of its new Hume Audio Decision API, a product that lets users query audio content directly. According to the announcement, users can ask questions such as how many speakers are present, whether anyone is laughing, or whether a caller is frustrated. The post frames the API as a flexible, question-driven tool for extracting decisions and insights from audio, positioning it around speaker detection and emotional-state analysis. No pricing, availability date, or technical benchmarks were disclosed in the post.
Hume AI releases audio-native API to detect caller emotion beyond transcripts
The post argues that voice agents and QA dashboards grade calls on transcripts, which discard tone, hesitation, and interruption cues, so angry callers can score green. It says speech-to-text and synthetic voices have advanced, but the agent in between still runs on text, and teams optimize what the transcript captures. Hume AI (@hume_ai) has released an API that answers plain-English questions about audio, such as whether a caller sounds unsure, returning a probability per answer. The API is audio-native, so it listens to the recording itself without a separate classifier per use case. Hume says it is trained on millions of human judgments and, in initial evals, matched a panel of human listeners more closely than a single human rater did, which the post says could let a lab score every model checkpoint instead of waiting on a panel.